Efficient multilingual keyword search using bloom filter for cloud computing applications

Efficient keyword search in electronic documents has been an important problem in computer science for the last many decades. With the popularity of cloud services, some applications require searching in multilingual environment. Other applications require data to be stored in the cloud in encrypted form and outsourced to a third party for processing. This paper proposes an algorithm using bloom filters to perform efficient multilingual search on data stored in the cloud in plain or encrypted form. When the user sends in a keyword to be searched, its language is first determined and its corresponding language list bloom filters are checked for presence of the keyword. To make the algorithm more efficient and accurate, we have created two categories of bloom filters namely primary and secondary bloom filter. The list of documents having the keyword is returned to the user. For secure applications, the encrypted documents and its corresponding bloom filters are stored in the server. When user wants to perform a search in stored encrypted documents it sends the keyword to the server. The server applies similar technique to return the encrypted documents having the keyword and the client uses the key to decrypt the documents if required. While searching for keywords, we test the word against the bloom filter of documents which enables these to be stored in encrypted form. Checking of a word against the bloom filter of its documents takes constant time. Experimental results show that searching for a word in encrypted documents can be performed quite efficiently using this scheme even if the environment is multilingual.